Plant cover as an estimator of above-ground biomass in semi-arid woody vegetation in Northeast Patagonia, Argentina

Abstract

The quantification of carbon storage in vegetation biomass is a crucial factor in the estimation and mitigation of CO2 emissions. Globally, arid and semi-arid regions are considered an important carbon sink. However, they have received limited attention and, therefore, it should be a priority to develop tools to quantify biomass at the local and regional scales. Individual plant variables, such as stem diameter and crown area, were reported to be good predictors of individual plant weight. Stand-level variables, such as plant cover and mean height, are also easy-to-measure estimators of above-ground biomass (AGB) in dry regions. In this study, we estimated the AGB in semi-arid woody vegetation in Northeast Patagonia, Argentina. We evaluated whether the AGB at the stand level can be estimated based on plant cover and to what extent the estimation accuracy can be improved by the inclusion of other field-measured structure variables. We also evaluated whether remote sensing technologies can be used to reliably estimate and map the regional mean biomass. For this purpose, we analyzed the relationships between field-measured woody vegetation structure variables and AGB as well as LANDSAT TM-derived variables. We obtained a model based ratio estimate of regional mean AGB and its standard error. Total plant cover allowed us to obtain a reliable estimation of local AGB, and no better fit was attained by the inclusion of other structure variables. The stand-level plant cover ranged between 18.7% and 95.2% and AGB between about 2.0 and 70.8 Mg/hm2. AGB based on total plant cover was well estimated from LANDSAT TM bands 2 and 3, which facilitated a model-based ratio estimate of the regional mean AGB (approximately 12.0 Mg/hm2) and its sampling error (about 30.0%). The mean AGB of woody vegetation can greatly contribute to carbon storage in semi-arid lands. Thus, plant cover estimation by remote sensing images could be used to obtain regional estimates and map biomass, as well as to assess and monitor the impact of land-use change on the carbon balance, for arid and semi-arid regions